ConstructVision: AI-driven Construction Progress Forecasting

ConstructVision is an AI-powered project that utilizes Convolutional Neural Networks (CNNs) to predict and track the progress of wall construction using real-time image data captured during the construction process.

Problem Statement

The aim of this project is to leverage deep learning techniques, specifically CNNs, to predict the progress at different stages of wall construction. This involves creating a comprehensive dataset, training a CNN model, and using it to analyze image frames for prediction accuracy.

Features

• Progress Prediction: Predicts the current stage of wall construction (e.g., foundation, 25% brickwork, 50% brickwork).
• Real-Time Image Processing: Uses OpenCV for real-time image preprocessing and predictions.
• Efficiency and Optimization: Incorporates fine-tuned pre-trained CNN models like VGG16 or ResNet50 for optimal results.

Steps

1. Dataset Creation

• Collect images of wall construction at various stages.
• Label each image according to the progress stage.
• Ensure diversity in the dataset to cover a wide variety of construction conditions.

2. Preprocessing of Images

• Resize images to a standardized size suitable for model input.
• Normalize pixel values to ensure consistency.
• Convert images to grayscale to simplify processing and reduce computational requirements.

3. Model Training

• Train the CNN model with labeled images as input and progress stages as output.
• Experiment with fine-tuning pre-trained networks (e.g., VGG16, ResNet50) or develop a custom architecture.
• Optimize the model for extracting features relevant to wall construction progress.

4. Image Processing Pipeline

• Use OpenCV to process captured images in real-time.
• Apply preprocessing steps like resizing and normalization.
• Pass images through the trained CNN model for progress prediction.

5. Progress Calculation

• Analyze predictions over time to calculate overall construction progress.
• Incorporate feedback loops to iteratively improve the accuracy of predictions.

Installation Instructions

1. Clone the repository:

git clone https://github.com/Afsheen-Aziz/ConstructVision.git

2. Navigate into the project directory:

cd ConstructVision

3. Install dependencies:

pip install -r requirements.txt

4. Start the application:

python app.py

Usage

1. Launch the local server by running the application.
2. Upload an image using the web interface.
3. View predicted progress and visualized results on the web interface.

Project Structure

ConstructVision/
├── app.py           # Main application logic
├── static/
│   └── uploads/    # Directory to store uploaded images
├── templates/
│   ├── index.html  # Webpage for file upload
│   └── result.html # Webpage for displaying results
├── weights/
│   ├── best.pt     # Trained model weights
│   └── last.pt     # Latest version of the model weights
└── README.md        # Project documentation

Contributing

We welcome contributions! To contribute:

1. Fork the repository.
2. Create a feature branch:

git checkout -b feature-name

3. Commit your changes:

git commit -m "Description of changes"

4. Push the branch:

git push origin feature-name

5. Submit a pull request.

License

This project is MIT licensed. See the LICENSE file for details.

Contact

For any queries, suggestions, or feedback:

• Name: Afsheen Aziz
• Email: afsheenonnar@gmail.com
• GitHub: https://github.com/Afsheen-Aziz

---

Thank you for exploring ConstructVision! Together, let's advance the future of AI in construction.